薄壁石墨鑄鐵之高溫熱疲勞性質之研究

Abstract

本研究的主要目的即是建立薄壁石墨鑄鐵之鑄造技術，鑄件之厚度標的為2~6mm，研究上探討一些相關製程及冶金參數對於顯微組織（包含球化率、縮化率、球墨數目、肥粒鐵、波來鐵、碳化物）的影響以及熱疲勞性質之關係，以期獲致優良薄壁石墨鑄鐵之最佳製程參數條件。 實驗結果顯示，球墨鑄鐵在固定3.0％C之條件下，球墨數目及球化率均隨Si含量之增加，首先增加，在約3.8~3.9%Si處達到最高值，然後再下降，肥粒鐵量則隨Si含量之增加而逐漸增加，在約4.6~4.8%Si時達到最高值，並維持一定值或稍降。又，在相同的C、Si含量下，2mm鑄件之球墨數目高於3mm鑄件，而球化率則差異不大。改變造模材料(化學模、濕砂模)，對於球墨數目與球化率之影響不大。針對澆鑄溫度對於球墨數目之影響而言，低溫澆鑄之鑄件其球墨數目高於高溫澆鑄者，其他條件下澆鑄溫度之影響並不顯著。對於基地組織而言，4.77%Si含量，可得到85.1%之肥粒鐵量，其3mm鑄件之肥粒鐵量則高於2mm鑄件，而化學模鑄件之肥粒鐵量則稍高於濕砂模鑄件，高溫澆鑄之鑄件肥粒鐵量亦稍高於低溫澆鑄。在相同Si含量下，影響肥粒鐵量以鑄件厚度最大，造模材料次之，澆鑄溫度則最小。 分析C、Si或C.E.對於熱疲勞性質之影響可得知，在過共晶組成，即C.E.值介於4.5~4.6%之間，並配合高Si低C(如：3.0%C+4.8%Si)組合，可以得到最佳之熱疲勞性質。最佳熱疲勞性質係對應於最高肥粒鐵量以及適當之球墨數目。此外，添加約0.5%Mo可以顯著提升熱疲勞壽命。綜合研究之實驗結果，對於薄件(2~3mm)球墨鑄鐵而言，最佳之合金設計為C：~3.0%、Si：4.7~4.8% (C.E.：4.5~4.6%)、Mo：0.5%。 試片在室溫~800℃之間進行熱循環時，於加熱過程中波來鐵及部份肥粒鐵(第一回)或部份肥粒鐵及麻田散鐵(第二回以後)會變態為沃斯田鐵，並於水淬火過程中變態為麻田散鐵。此外，在每次之熱循環加熱過程中，未變態之麻田散鐵會被加熱而形成回火麻田散鐵。隨著熱循環次數的增加，麻田散鐵在高溫回火過程中會逐漸分解而析出二次石墨於晶界上，使得基地之含C量下降，致使在後續之熱循環變態過程中，肥粒鐵量逐漸增加，而麻田散鐵及回火麻田散鐵則逐漸降低。在熱循環過程中，由於晶界上逐漸析出二次石墨，且受到每一回熱循環之拉應力作用下，在晶界上之二次石墨周圍會發生裂紋，而熱循環之主裂紋的生長即沿著晶界，連結這些微小裂紋而逐漸擴展，直至斷裂為止。 鑄件厚度為2-3mm時之薄壁縮墨鑄鐵，大都仍以球狀石墨為主(縮化率偏低)，熱疲勞壽命最高為添Mo的球墨鑄鐵>添加Mo的縮墨鑄鐵>球墨鑄鐵>縮墨鑄鐵。當鑄件厚度為6mm時，薄壁縮墨鑄鐵能獲得縮化率>65%的縮墨鑄鐵，熱疲勞壽命最高為添加Mo的縮墨鑄鐵>縮墨鑄鐵(縮化率>70%)≒球墨鑄鐵>縮墨鑄鐵(縮化率<60%)>片墨鑄鐵。試片中央熱膨脹變形量最高的為球墨鑄鐵及縮墨鑄鐵，其次為添加Mo的縮墨鑄鐵，熱膨脹變形量最低的為片墨鑄鐵。在考量熱疲勞壽命及熱膨脹變形量的因素下，可選擇添加Mo的縮狀石墨鑄鐵。

The primary purpose of this research is to establish the optimal casting conditions for producing thin-section (2~6mm) spheroidal and compected graphitic cast irons for high temperature applications (up to 800oC). Experimentally, the microstructures (include nodularity, vermicularity, nodule count, %ferrite, %pearlite, and %carbide) and thermal fatigue property will be evaluated and correlated with alloy design and casting parameters, such as molding material and pouring temperature. The results show that, for a fixed C content of some 3%, both the nodule count and nodularity increase first with increasing Si content, reach maxima at around 3.8~3.9%Si, and then decrease with further increase in Si content. On the other hand, the percent ferrite increases gradually with increasing Si content, reaches the maximum at around 4.6~4.8%Si, and then remains more or less constant or increases slightly. Regarding the effects of casting parameters on microstructure, the results show that higher nodule counts can be obtained in castings with a thinner section or with a lower pouring temperature. However, the effect of molding material (chemically-bonded sand and green sand) on nodule count is not significant. On the other hand, all the above three casting parameters exert little influence on graphite nodularity. Furthermore, the percent ferrite is higher in castings with a thicker section, a higher pouring temperature and molded with chemically-bonded sand. The optimal alloy design for attaining the best thermal fatigue property has been found to be a slightly hypereutectic composition, i.e., 4.5~4.6%CE, with a combination of relatively low C content and high Si content, e.g., 3.0%C + 4.8%Si. In addition, the best thermal fatigue property corresponds to a microstructure with the highest ferrite content and moderate nodule count. Furthermore, adding some 0.5%Mo to the iron significantly increases the thermal fatigue life. Based upon the results obtained herein, the optimal alloy design is: C: ~3%, Si: 4.7~4.8% (CE: 4.5~4.6%), Mo: 0.5%. During the thermal fatigue test, all the pearlite and part of the ferrite in the as-cast condition transform to austenite during the heating stage of the first cycle, and the austenite will transform to martensite after cooling (quench) to room temperature. As a result, un-transformed ferrite and martensite are present in the microstructure after one cycle. In the second cycle, part of un-transformed ferrite and martensite will again transform to austenite during the heating stage, while at the same time the un-transformed martensite will be tempered. Consequently, ferrite, martensite and temper martensite will be present in the microstructure after the specimen being cooled to room temperature. The afore-mentioned transformation mechanism continues to operate during the subsequent thermal fatigue cycles. However, at some point, the repeated tempering of both martensite and temper martensite causes the precipitation of secondary graphite particles at the grain boundaries, gradually reducing the dissolved carbon content in the matrix. As a result, the volume fractions of both martensite and temper martensite decrease gradually, while the ferrite and secondary graphite increase, with the progress of the thermal fatigue cycle. Under the influence of the tensile stress during each thermal fatigue cycle, cracks start to initiate at the graphite-matrix interface and/or at the vicinity of the precipitated secondary graphite particles. And then, cracks propagate along the grain boundary until the occurrence of complete fracture. When the compacted graphite cast iron is the graphite type that thin wall of the 2-3 mm, mostly still regarding the spheroidal graphite as principle. The cast irons with adding some 0.5%Mo of the spheroidal graphite exhibit the best thermal fatigue life. When the compacted graphite cast iron is the graphite type acceptable compacted graphite structure can only be obtained when the section thickness exceeds about 6-mm. The best thermal fatigue property is adding some 0.5%Mo of the compacted graphite cast iron. Regarding the constrained cyclic thermal fatigue property (RT ~ 800oC), the extent of swelling at the thermal center of the test specimens is highest for irons with spheroidal graphite, followed by irons with compacted graphite, and then CG irons with Mo addition, and the irons with flake graphite is the least. cast irons with spheroidal graphite exhibit the best thermal fatigue life, which is followed by irons with compacted graphite, and then flake graphite cast irons.